You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: content/posts/newsletter-008/index.md
+70-32Lines changed: 70 additions & 32 deletions
Original file line number
Diff line number
Diff line change
@@ -83,36 +83,70 @@ Highlights of [this month's update][will_blogpost] include:
83
83
84
84
### gfx-rs and wgpu news
85
85
86
-
[gfx-hal-0.5](https://github.com/gfx-rs/gfx/) was released:tada:
86
+
[gfx-hal-0.5](https://github.com/gfx-rs/gfx/) was released!
87
87
Improvements done in March:
88
-
- Debug markers. Users are now able to debug-annotate parts of the rendered frame, so that inspecting it in a GPU debugger is more enjoyable.
89
-
- The generic range parameters are removed in favor of simple structs. This is a move towards simpler low-level API.
90
-
- Physical device features for NDC Y-flip and sampler mirror clamp are added
91
-
- Physical device performance hints are introduced. The first hint is for "base vertex/instance" support.
92
-
-`SmallVec` is removed from the API, it's reshaped to avoid any heap allocations. Previously, it had to touch the heap on multiple descriptor sets or command buffers.
93
-
- DX12 got true support for read-only storage bindings. This is one of the opt-in derivations from Vulkan that allow to better map users logic to non-Vulkan backends, also used by WebGPU.
94
-
- Last but not the least, @zicklag[has been fighting](https://github.com/gfx-rs/gfx/pull/3151) with the OpenGL backend to align its API with the rest of the crowd, armed with [surfman](https://github.com/pcwalton/surfman). The fight is reading conclusion, and we are crossing fingers to add OpenGL support to `wgpu-rs` as it lands.
95
-
96
-
[wgpu](https://github.com/gfx-rs/wgpu) and [wgpu-rs](https://github.com/gfx-rs/wgpu-rs) changes in March:
97
-
-@grovesNL reached an epic milestone in the Web target by showing the [first triangle](https://github.com/gfx-rs/wgpu-rs/pull/193#issuecomment-599156540). Users will soon be able to seamlessly target the web with their existing `wgpu-rs` applications :rocket:
98
-
-`wgpu-types` crate is created to share types between the Web target and the native one.
99
-
-@lachlansneff improved the _async_ story quite a bit, we also converted more methods to be asynchronous.
100
-
- Debug labels support.
101
-
- Id management story for browsers with a GPU process has been completely redesigned and now working well.
102
-
- All the objects are properly destroyed and GPU tracked if needed.
103
-
- Ability to provide a `Surface` so that the selected adapter can present to it.
104
-
- New "mailbox" present mode.
88
+
89
+
- Debug markers.
90
+
Users are now able to debug-annotate parts of the rendered frame,
91
+
so that inspecting it in a GPU debugger is more enjoyable.
92
+
- The generic range parameters are removed in favor of simple structs.
93
+
This is a move towards simpler low-level API.
94
+
- Physical device features for NDC Y-flip and sampler mirror clamp are added.
95
+
- Physical device performance hints are introduced.
96
+
The first hint is for "base vertex/instance" support.
97
+
-`SmallVec` is removed from the API, it's reshaped
98
+
to avoid any heap allocations.
99
+
Previously, it had to touch the heap on multiple descriptor sets
100
+
or command buffers.
101
+
- DX12 got true support for read-only storage bindings.
102
+
This is one of the opt-in derivations from Vulkan that allow to better map
103
+
users logic to non-Vulkan backends, also used by WebGPU.
104
+
- Last but not the least, @zicklag
105
+
[has been fighting](https://github.com/gfx-rs/gfx/pull/3151)
106
+
with the OpenGL backend to align its API with the rest of the crowd,
107
+
armed with [surfman](https://github.com/pcwalton/surfman).
108
+
The fight is reading conclusion, and we are crossing fingers
109
+
to add OpenGL support to `wgpu-rs` as it lands.
110
+
111
+
[wgpu](https://github.com/gfx-rs/wgpu) and
112
+
[wgpu-rs](https://github.com/gfx-rs/wgpu-rs) changes in March:
113
+
114
+
-@grovesNL reached an epic milestone in the Web target
115
+
by showing the [first triangle][wgpu-web-triangle].
116
+
Users will soon be able to seamlessly target the web
117
+
with their existing `wgpu-rs` applications. 🚀
118
+
-`wgpu-types` crate is created to share types between the Web target
119
+
and the native one.
120
+
-@lachlansneff improved the _async_ story quite a bit,
121
+
we also converted more methods to be asynchronous.
122
+
- Debug labels support.
123
+
- Id management story for browsers with a GPU process
124
+
has been completely redesigned and now working well.
125
+
- All the objects are properly destroyed and GPU tracked if needed.
126
+
- Ability to provide a `Surface` so that the selected adapter can present to it.
127
+
- New "mailbox" present mode.
105
128
106
129
Satellite projects:
107
-
-[naga](https://github.com/gfx-rs/naga) - the new in-house shader translator has reached the milestone of successfully loading a WGSL [boids example](https://github.com/gfx-rs/naga/blob/thda1f6a47b06c89abb1dff70326c076f1088964a3/test-data/boids.wgsl) and generating a valid Metal source for it :tada:
108
-
-[metal-rs](https://github.com/gfx-rs/metal-rs/) has got a lot of contribution by @adamnemecek. Indirect command encoding is particularly exciting!
109
-
-[gfx-extras](https://github.com/gfx-rs/gfx-extras) is a new library that is forked from rendy-memory/descriptor.
110
-
-[gfx-ocean](https://github.com/gfx-rs/gfx-ocean) was moved to gfx-rs organization and updated to gfx-hal-0.5.
111
-
-[gfx-portability](https://github.com/gfx-rs/portability) was also updated.
130
+
131
+
-[naga] - the new in-house shader translator has reached the milestone
132
+
of successfully loading a WGSL [boids example]
133
+
and generating a valid Metal source for it. 🎉
134
+
-[metal-rs](https://github.com/gfx-rs/metal-rs)
135
+
has got a lot of contribution by @adamnemecek.
136
+
Indirect command encoding is particularly exciting!
137
+
-[gfx-extras](https://github.com/gfx-rs/gfx-extras) is
138
+
a new library that is forked from rendy-memory/descriptor.
139
+
-[gfx-ocean](https://github.com/gfx-rs/gfx-ocean) was moved
140
+
to gfx-rs organization and updated to gfx-hal-0.5.
141
+
-[gfx-portability](https://github.com/gfx-rs/portability) was also updated.
@@ -287,7 +322,9 @@ or [join the next meeting][join].
287
322
-[Amethyst's "good first issue" issues][amethyst-issues];
288
323
-[A/B Street's "good first issue" issues][abstreet-issues];
289
324
-[Mun's "good first issue" issues][mun-issues];
290
-
- Anybody wants to work on the [GLSL front-end](https://github.com/gfx-rs/naga/issues/23) in Naga? One day, we'll be able to finally replace glsl-to-spirv, which is used by a lot of graphics applications and is prone to issues.
325
+
- Anybody wants to work on the [GLSL front-end][naga-issue] in Naga?
326
+
One day, we'll be able to finally replace glsl-to-spirv,
327
+
which is used by a lot of graphics applications and is prone to issues.
0 commit comments